| HOME | | | Japanese | | | Profile and Publication List |
Language: English | Japanese
Angew. Chem. Int. Ed. 2012, 51, 7263‒7266
| Highly Diastereoselective Construction of Acyclic Systems with Two Adjacent Quaternary Stereocenters by Allylation of Ketones Takeshi Takeda, Masanori Yamamoto, Satoshi Yoshida, Akira Tsubouchi Angew. Chem. Int. Ed. 2012, 51, 7263‒7266 DOI: 10.1002/anie.201202808 [pdf] |
 |
Abstract: The use of titanocene (Cp2Ti) as a Lewis acid in allylation of asymmetric ketones enables us to achieve the construction of two adjacent quaternary stereocenters in a highly efficient and highly diastereospecific manner. In this reaction, nucleophilic attack to ketones starting from (E)-allylic sulfides gives anti-homoallylic alcohols, while the reaction from (Z)-allylic sulfides affords syn-homoallylic alcohols with excellent diastereoselectivities. Several other findings including the effect of irradiation, temperature, and solvents are also discussed in the commentary. |
Introduction
Stereoselective construction of carbon-carbon bonds in regio-selective, chemo-selective, and stereoselective manners is a general important theme in organic synthesis. Among them, the formation of consecutive quaternary stereocenters in acyclic system has been one of the most challenging tasks, especially in case of all-carbon systems.1 In natural systems, such a consecutive quaternary stereocenters is synthetically achieved via cyclization or cation-transfer reaction of poly-enes, and biomimetic cascade reactions have been investigated for the purpose (Scheme 1).1c
Scheme 1 | Stereoselective construction of consecutive quaternary stereocenters by E. J . Corey and co-workers.1c
These examples are rather case-sensitive with limited substrate scopes. For more general ways with a wider substrate scope, face-selective nucleophilic attack of carbon reagents to carbon-heteroatom double-bonds has been investigated, and aldol reactions are known for stereoselective synthesis of tertiary alcohols.2 However, the construction of sequential quaternary carbons using α-substituted carbonyl compounds requires the precise control of stereochemistry of the corresponding enolates, and this remains challenging..
Scheme 2 | Construction of consecutive quaternary stereocenters by aldol reactions using cuprous reagents.
Nucleophilic attack of 3,3-di-substituted allylic metal species toward carbon-heteroatom double-bonds has also been investigated for the purpose.3 In this category, 3-substituted-2-cyclohexenyl metals have been well studied,4 while the use of aliphatic, linear 3,3-disubstituted allylic metal species remains challenging owing to the difficulty in controlling the stereochemistry or poor reactivity by their steric hindrance. Prof. Takeshi Takeda (Tokyo University of Agriculture and Technology, Japan) reported a high reactivity of allylic titanocene analogues in these reactions, and allylic titanocenes derived from 3-substituted allylic sulfide were used to attack to asymmetric ketones, leading to the corresponding anti-homoallylic alcohols with a good diastereoselectivity irrespective of the E/Z ratio in the original allylic sulfides (Scheme 3).5
Scheme 3 | Reaction between a linear 3,3-disubstituted allylic derivative and electrophiles.
The formation of (E)-allylic titanocenes is thermodynamically favored, and the reaction of the (E)-titanocenes with electrophilic ketones occurs via the six-membered cyclic chair-like transition state with bulkier substituent of ketones being at the pseudo-equatorial (TS1), and this rationalizes the high diastereoselectivity realized even with aliphatic ketones such as 2-octanone (Figure 1).
Figure 1 | Zimmerman‒Traxler Chair-Like Transition States.
This reaction can be extended to γ-di-substituted allylic titanocenes with various asymmetric ketones. In this case, construction of consecutive adjacent quaternary carbon centers can be achieved in the anti-selective manner. In this work, we investigated the nucleophilic attack reaction of 3,3-di-substituted allylic titanocenes derived from the corresponding allylic alcohols to construct two adjacent quaternary carbon stereocenters. We found that the reaction occurs in a good yield with excellent stereoselectivity in a "stereospecific" manner with various ketones in terms of the bulkiness. The isolate yield reaches 92%, and the diastereoselectivity is up to 99:1. The followings are the details of this work.
Initial Screening
As previously discussed, 3-substituted allylic sulfides give thermodynamically favored E-allylic titanocenes, and they react with electrophilic ketones to give the anti-homoallylic alcohols selectively via the six-membered chair-like transition states (Figure 1). On the contrary, allylic titanocenes from 3,3-disubstituted allylic sulfide 1a and 3,3-di-substituted allylic chloride 1b did not give a good diastereoselectivity (Table 1). The difference in the stability of 3,3-di-substituted allylic titanocenes for (E)-isomer and (Z)-isomer may be small, and therefore, the selective formation of the (E)-allylic titanocenes may be difficult. In this case, the mixture of E/Z-allylic titanocenes would be formed, and the subsequent reaction with ketones will give the corresponding mixture of the products of homoallylic alcohols.
By the way, we recognized that the allylic chloride 1b shows the temperature-dependence of the stereoselectivity, while the allylic sulfide 1a does not show such a temperature dependence (Table 1). As was discussed earlier, the control of the stereochemistry of allylic metal species plays a crucial role in controlling the stereoselectivity of the final products. As such, we then prepared (E)- and (Z)-allylic alcohol derivatives with higher regiopurity, and they were then reduced by Cp2TiII to give the allylic titanocenes. We selected geraniol and nerol as the (E)- and (Z)-3,3-di-substituted allylic alcohols, and the corresponding geranyl sulfide E-1c and neryl sulfide Z-1c were used as starting materials (Table 2).
As a result, the stereoselectivity somehow reflects the E/Z ratio of the allylic alcohols. Especially, (Z)-neryl sulfide Z-1c showed the different tendency with another stereoisomer being the major product at lower temperature (entry 3). This indicates the temperature dependent isomerization of allylic metal species.6 Its suppression will lead to the stereospecific reaction. Therefore, we then used the geranyl sulfide E-1c with the E/Z ratio determined. We found the relatively strong relationship between the reaction temperature and the stereoselectivity (Table 3, entry 3‒5).
However, some results were deviated from the hypothesis "the lower the reaction temperature is, the higher the selectivity is" (entry 6). This may indicate that other factors also govern the stereoselectivity. Allylic compounds are known to isomerize upon irradiation, and therefore, we conducted the reactions in the dark, and a stark contrast was recognized (Table 4, entries 3 & 4). With this in mind, the following investigations were all conducted in the dark.
Thus, we concluded that the following conditions are indispensable for achieving high diastereoselectivity in nucleophilic attack of allylic titanocenes to ketone electrophiles, when 3,3-di-substituted allylic sulfides are used as the starting building block:
A) The oxidative addition (reduction of allylic sulfides by Cp2TiII) should be conducted at low temperatures (~‒30 ℃)
B) Dark condition is better for suppressing the photoisomerization of allylic titanocenes.
C) High E/Z ratio, which is reflected to the final diastereoselectivity.
The (E)-allylic sulfides give the anti-products, while the (Z)-allylic sulfides afford the corresponding syn homoallylic alcohols as the products. Therefore, we concluded that the stereospecific reaction takes place under the conditions mentioned above (Scheme 4).
Scheme 4 | Proposed mechanism.
By the way, it is reported that the oxidative addition at temperatures lower than ‒50 ℃ does not proceed,7 and therefore, we never tried to conduct the reactions at temperatures lower than ‒45 ℃. Similarly, the use of allylic ethers and allylic esters may also show such a stereospecific behavior with a high stereoselectivity, but they require higher reaction temperatures at the initial oxidative addition via reduction of allylic ethers/esters by Cp2TiII.5a,8 Because the lower reaction temperatures seem to be indispensable for the better stereoselectivity, we did not investigate ethers and esters as the leaving groups in the initial oxidative addition. We also noticed that the reaction was almost accomplished within 4 h in the second step (entry 2 in Table 3).
Solvent Effect
We sometimes recognize a significant effect of solvents in chemical yield and selectivity.9,10 As such, we screened several solvents including hexane, dioxane, tetrahydrofuran (THF), and diethyl ether (Et2O). Initial screening showed that the use of ether is better for higher isolated yield (Table 5), and therefore, we continued to investigate other solvents as well.
As a result, we found that the use of cyclopentyl methyl ether (CPME) as the initial preparation of low-valence titanium Cp2TiII species from Cp2TiIVCl2 and n-butyl lithium, while THF is better as the solvent for the subsequent addition of allylic sulfides and ketones (entries 5 & 9). On the contrary, the use of other solvents or the use of CPME only do not give high yield (entries 6‒8). Unlike the isolated yield, the stereoselectivity is less affected by the choice of solvents.
Substrate Scope
To check the generality of this reaction, we then conducted the reaction using various allylic sulfides as the precursors of allylic titanocenes and various ketones as the electrophiles (Table 6). When an acetophenone derivative 2c and cyclohexyl methyl ketone 2d are used, the high diastereoselectivity is achieved. The use of propiophenone 2g also gives a good selectivity which is higher than those reported by Prof. Iran Marek and co-workers.11 However, 4-phenyl-2-butanone 2e and methyl ethyl ketone 2f do not give a satisfying result (entries 3, 4, 7, 8).
As previously discussed, high diastereoselectivity requires the control of the stereochemistry in allylic titanocenes. In addition, the reaction should be took place via the six-membered chair-like transition state with the bulkier substituent of ketones being at the pseudo-equatorial position (Figure 1, TS1). In this regard, the difference of Ea‡ between TS1 and TS2 will be very small when methyl ethyl ketone 2f with the smallest difference in the bulkiness of the substituents is used. Quantitatively, the use of the diastereoselectivity as the provisional rate constant and the assumption that the pre-exponential factors in the Arrhenius equation are similar, we can estimate the energy of the transition states.12,13
From this model, the energy difference for the transition states, ΔEa‡, at ‒78 ℃ was estimated to be 2.7 kJ mol‒1 for the (E)-allylic titanocene and 4-phenyl-2-butanone 2e using the diastereoselectivity (dr = 81:19) in entry 7 of Table 6 as the rate constants on equation 1. Provided that the energy is less dependent on the temperature within a narrow range of temperature, improved selectivity of 89:11 is expected for the reaction ‒120 ℃ (153 K). In fact, the addition of the allylic titanocene to the methyl ethyl ketone at the temperature of ‒110 ℃ using an ethanol/nitrogen slush bath.
Conclusion
Thus, the sterically hindered titanocene enables us to achieve excellent stereoselectivity with asymmetric dialkyl ketones as well as acetophenone derivatives. Besides, the E-Z isomerization is suppressed for 3,3-di-substituted allylic titanocenes under optimized conditions, and the subsequent nucleophilic addition to ketones achieves the stereospecific construction of two adjacent quaternary stereocenters. This work was conducted as an undergraduate study by Masanori Yamamoto at Tokyo University of Technology and Agriculture, and the results have been published in Angewandte Chemie International Edition.14,15
Last modified on 11th May 2023
Contact Information
|
Assistant Professor of Yamanaka Lab. |
References and Notes
| 1. | (a) E. A. Peterson, L. E. Overman, Proc. Nat. Acad. Sci. USA 2004, 101, 11943‒11948; (b) 野依良治ら編, 大学院講義有機化学Ⅱ, 1998, 1; (c) E. J. Corey, S. Lin, J. Am. Chem. Soc. 1996, 118, 8765‒8766. |
| 2. | (a) J. Deschamp, O. Chuzel, J. Hannedouche, O. Riant, Angew. Chem. Int. Ed. 2006, 45, 1292‒1297; (b) F. Douelle, A. S. Capes, M. F. Greaney, Org. Lett. 2007, 1931‒1934. |
| 3. | (a) L. R. Reddy, B. Hu, M. Prashad, K. Prasad, Org. Lett., 2008, 10, 3109‒3112; (b) H. Ren, G. Dunet, P. Mayer, P. Knochel, J. Am. Chem. Soc., 2007, 129, 5376‒5377; (c) aluminium; Z. Peng, T. D. Blumke, P. Mayer, P. Knochel, Angew. Chem. Int. Ed. 2010, 49, 8516‒8519. |
| 4. | (a) S. Araki, M. Hatano, H. Ito, Y. Butsugan, J. Organomet. Chem. 1987, 329‒375; (b) G. Dunet, P. Mayer, P. Knochel, Org. Lett. 2008, 10, 117‒120. |
| 5. | (a) Y. Yatsumonji, T. Nishimura, A. Tsubouchi, K. Noguchi, T. Takeda, Chem. Eur. J. 2009, 15, 2680‒2686; (b) T. Takeda, H. Wasa, A. Tsubouchi, Tetrahedron Lett. 2011, 4575‒4578. |
| 6. | A. Yanagisawa, S. Habaue, K. Yasue, H. Yamamoto, J. Am. Chem. Soc. 1994, 116, 6130‒6141. |
| 7. | (a) J. X. McDermott, G. M. Whiteside, J. Am. Chem. Soc. 1974, 96, 947‒948; (b) J. X. McDermott, M. E. Wilson, G. M. Whiteside, J. Am. Chem. Soc. 1976, 98, 6529‒6536. |
| 8. | Y. Yatsumonji, T. Sugita, A. Tsubouchi, T. Takeda, Org. Lett. 2010, 12, 1968‒1971. |
| 9. | 新井健、妹尾学、浅原照三、有機化学反応における溶媒効果、産業図書、1970 |
| 10. | R. Hirabayashi, C. Ogawa, M. Sugiura, S. Kobayashi, J. Am. Chem. Soc. 2001, 123, 9493‒9499. |
| 11. | B. Dutta, N. Gilboa, I. Marek, J. Am. Chem. Soc. 2010, 132, 5588‒5589. |
| 12. | Both the theoretical and experimental results have shown these tendencies. For theoretical calculation of free energies for the transition states with allyl silanes, see; L. F. Tietze, T. Kinzel, S. Schmatz, J. Am. Chem. Soc. 2006, 128, 11483‒11495. |
| 13. | Later the author recognized that this assumption is not suitable. Nevertheless, it is noteworthy that the high reactivity of titanocenes which is not damaged at such a low temperature enables us to achieve the good selectivity even with aliphatic alkyl ketones. |
| 14. | T. Takeda, M. Yamamoto, S. Yoshida, A. Tsubouchi, Angew. Chem. Int. Ed. 2012, 51, 7263‒7266. |
| 15. | Reference [14] is introduced in Synfacts. |